OpenK3
Get Started
Menu
FeaturesDocsBlogKimi K3Kimi CodeKimi K2.7 CodeAboutContact

Kimi K2.7 Code guide

Kimi K2.7 Code: Official Specs, API & Pricing

Kimi K2.7 Code is Moonshot AI's specialized coding model. It is served through the Kimi API under the model IDs kimi-k2.7-code and kimi-k2.7-code-highspeed, with a 262,144-token (256K) context window and always-on thinking. As of July 24, 2026 it appeared in the official Kimi API model list alongside Kimi K3, with lower official CNY per-token pricing than K3 across all billing categories. [K3, K5, K7]

Official K2.7 Code QuickstartK3 vs K2.7 CodeKimi K3 Guide

Overview

What Kimi K2.7 Code is

Kimi K2.7 Code is positioned as a specialized coding model, in contrast to Kimi K3, which Moonshot AI positions as its flagship model for long-horizon coding, knowledge work, and deep reasoning. K2.7 Code remains available through the Kimi API after the K3 release. [K3, K1]

The official sources reviewed for this page did not state that Kimi K3 replaces Kimi K2.7 Code. Both models appeared in the official Kimi API model list as of July 24, 2026. [K5]

Key features

Key Features of Kimi K2.7 Code

All capability descriptions are based on official Kimi API documentation, each with a dated access check.

Specialized coding model

Kimi K2.7 Code is positioned as a specialized coding model in the official Kimi API documentation and model list. [K3, K5]

256K context window

Kimi K2.7 Code supports a context window of 262,144 tokens (256K). [K3, K4, K7]

Always-on thinking

Thinking is always on. Kimi K2.7 Code uses thinking: {"type": "enabled"}; no reasoning-intensity control is documented. [K3, K4]

Highspeed variant

A kimi-k2.7-code-highspeed variant is listed in the official model list. Moonshot AI states it outputs approximately 180 tokens per second, reaching up to 260 tokens per second in short-context scenarios. Independent speed verification has not been performed. [K5]

Tool calling

Kimi K2.7 Code supports tool calling with tool_choice values of auto and none. The required value is not supported and returns an error. [K3, K4]

Automatic context caching

Context caching is automatic, stated in the pricing page feature summary. A cache hit rate was not stated in the official sources accessed. [K7]

API reference

API model IDs and parameters

These fields come from the official Kimi API documentation and may change over time.

Kimi K2.7 Code API model IDs and parameters
FieldValueSources
API model IDkimi-k2.7-code; high-speed variant: kimi-k2.7-code-highspeedK3, K5
Context window262,144 tokens (256K)K3, K4, K7
Thinking / reasoning controlThinking is always on via thinking: {"type": "enabled"}; no intensity control is documented.K3, K4
Default max output tokens32,768. The maximum configurable value was not confirmed in the accessed official sources as of July 24, 2026.K3
tool_choiceSupports auto and none only; required is not supported and returns an error.K3, K4
Dynamic tool loadingNot documented in the K2.7 Code quickstart as of July 24, 2026; absence from that guide does not confirm unsupported status.K3
Structured outputThe K2.7 Code pricing page lists JSON Mode; detailed interface documentation was not found in the quickstart as of July 24, 2026.K7
Partial ModeListed in the K2.7 Code pricing page feature summary; detailed usage documentation was not found in the quickstart as of July 24, 2026.K7
Context cachingAutomatic, stated in the pricing page feature summary. A cache hit rate was not stated in the official sources accessed.K7

API pricing

Kimi K2.7 Code CNY pricing per 1M tokens

Pricing below comes from the official Kimi K2.7 Code CNY pricing page, accessed July 24, 2026. USD pricing was not stated in the official sources accessed. Prices may change; check the current official pricing page for the latest rates.

Kimi K2.7 Code CNY pricing summary
Billing CategoryCNY per 1M tokens
Input — cache hit¥1.30 standard; ¥2.60 Highspeed
Input — cache miss¥6.50 standard; ¥13.00 Highspeed
Output¥27.00 standard; ¥54.00 Highspeed

CNY pricing is documented on the Kimi K2.7 Code pricing page. USD pricing was not stated in the official sources accessed as of July 24, 2026. [K7]

K2.7 Code and K3

How K2.7 Code relates to Kimi K3

Both kimi-k3 and kimi-k2.7-code, plus kimi-k2.7-code-highspeed, appear in the official Kimi API model list as of July 24, 2026. [K5]

The official sources reviewed for this page did not state that Kimi K3 replaces Kimi K2.7 Code. [K5]

For a side-by-side breakdown, see the Kimi K3 vs K2.7 Code comparison and the Kimi K3 guide.

Known limitations

What remains unresolved

Kimi K2.7 Code known limitations and unknowns
LimitationDetail
Architecture and parameter count not statedThe official sources accessed did not publish Kimi K2.7 Code's architecture, total parameters, or active parameters.
USD pricing not statedKimi K2.7 Code USD pricing was not stated in the official sources accessed as of July 24, 2026.
Open weights status not coveredThis page covers the hosted Kimi K2.7 Code API. An official open-weight release for K2.7 Code was not confirmed in the sources accessed.
Vendor-reported speed not independently verifiedThe Highspeed throughput figures (~180 tokens/sec, up to ~260 tokens/sec) are vendor-stated; OpenK3 has not independently verified them.
No independent benchmark testingOpenK3 has not executed original coding benchmarks comparing K2.7 Code to other models.

FAQ

Common questions about Kimi K2.7 Code

What is Kimi K2.7 Code?

Kimi K2.7 Code is Moonshot AI's specialized coding model, served through the Kimi API under the model IDs kimi-k2.7-code and kimi-k2.7-code-highspeed, with a 262,144-token (256K) context window and always-on thinking. [K3, K5]

What is the Kimi K2.7 Code API model ID?

The standard model ID is kimi-k2.7-code, and the high-speed variant is kimi-k2.7-code-highspeed, as listed in the official Kimi API model list. [K3, K5]

What is the Kimi K2.7 Code context window?

Kimi K2.7 Code supports a context window of 262,144 tokens (256K), based on official documentation accessed July 24, 2026. [K3, K4, K7]

How much does the Kimi K2.7 Code API cost?

As of July 24, 2026, official CNY pricing per 1M tokens is ¥1.30 cache-hit input (¥2.60 Highspeed), ¥6.50 cache-miss input (¥13.00 Highspeed), and ¥27.00 output (¥54.00 Highspeed). USD pricing was not stated in the official sources accessed. Prices may change. [K7]

Does Kimi K3 replace Kimi K2.7 Code?

No official source accessed as of July 24, 2026 says Kimi K3 replaces Kimi K2.7 Code. Both kimi-k3 and kimi-k2.7-code appeared in the official Kimi API model list on the access date. [K5]

What is the difference between K2.7 Code and K2.7 Code Highspeed?

Both are listed in the official model list. Moonshot AI states the Highspeed variant outputs approximately 180 tokens per second, reaching up to 260 tokens per second in short-context scenarios, and it has higher official CNY per-token pricing than the standard variant. Independent speed verification has not been performed. [K5, K7]

Related

Official sources

Documents used for this guide

Each source below includes its own access date.

  1. [K1]
    Kimi K3: Open Frontier Intelligence — Moonshot AI official blogAccessed July 24, 2026https://www.kimi.com/blog/kimi-k3
  2. [K3]
    Kimi K2.7 Code quickstart documentationAccessed July 24, 2026https://platform.kimi.com/docs/guide/kimi-k2-7-code-quickstart
  3. [K4]
    Kimi API model parameter referenceAccessed July 24, 2026https://platform.kimi.com/docs/api/models-overview
  4. [K5]
    Kimi official model listAccessed July 24, 2026https://platform.kimi.com/docs/models
  5. [K7]
    Kimi K2.7 Code CNY pricing pageAccessed July 24, 2026https://platform.kimi.com/docs/pricing/chat-k27-code